The statements in this section merely provide background information related to the present disclosure and may not constitute the prior art.
Standardized technologies for compressing image data currently include H.261, H.263, H.264, MPEG-2, MPEG-4, and so on. According to such conventional technologies for compressing an image, each image is divided into macroblocks having a fixed size and then encoded in the unit of macroblocks.
When the size of an input image is not an integer multiple of the size of a macroblock, the input image is padded such that it becomes an integer multiple of the size of the macroblock, all luma components and all chroma components of each macroblock within the padded input image are spatially or temporally predicted, and then a predicted residual is compressed through processes such as a transform, a quantization, and an entropy coding.
However, in a general technology for compressing an image, when the size of the input image is not an integer multiple of the size of the macroblock, it is difficult to determine an optimal type of the macroblock and an optimal prediction mode because the macroblock type and prediction mode are determined including a padded area after the input image is padded such that the size of the input image becomes an integer multiple of the size of the macroblock, and the encoding efficiency may be deteriorated because a bitstream including data of the padded area is generated.
Further, since decoding an image is carried out in correspondence with the image encoding, the already decreased compression efficiency of encoding a high resolution image would make it difficult to expect a highly efficient decoding.